Multiscale Modeling Approach to Understanding PKA Activation in Cardiac Myocytes

The [Beta]-adrenergic pathway in cardiac myocytes activates protein kinase A (PKA) to phosphoregulate several Ca²⁺ handling proteins, including the L-type Ca channel, ryanodine receptor, and sarco/endoplasmic reticulum Ca²⁺-ATPase (SERCA) (via phospholamban), resulting in inotropic, lusitropic and chronotropic responses. However, based on PKA kinetics measured in vitro, almost all PKA should be activated by basal concentrations of cAMP. Several recent studies have postulated a role for local degradation of cAMP by phosphodiesterases (PDE) in maintaining microdomains with lower cAMP concentrations around A kinase anchoring proteins (AKAP) which bind both PKA and PDE. In this study, we began by exploring if local clustering of AC and PDE combined with geometric constraints based on TEM images is sufficient to maintain a local difference in cAMP concentration. We show that even in a physiological geometry with significant steric hindrance and known PDE and AC kinetics, a non-physiological amount of PDE would be required to create a significantly difference cAMP concentration inside the cleft. Only when the diffusion coefficient for cAMP was lowered to 10 [Mu]m²/s or less would there be enough PDE in the cell to locally regulate the cAMP concentration and even then there would only be enough to protect the clefts and not any of the other compartments in the cell. We also explored how various metal ions may affect PKA activity. We show that while metal is essential to PKA catalysis, an overabundance decreases the rate of PKA activity. Through computational models, we determined that this biphasic relationship with metal could be explained by a second metal ion acting as a "linchpin" preventing ADP release until after the first metal ion leaves. Furthermore, the kinetics of this interaction were shown to be very metal ion dependent, suggesting that the concentrations of various metal ions in the cell will play a role PKA activation. While this helped to explain PKA activity after the regulatory (R) and catalytic (C) subunits release, questions remained about how PKA activity is regulated by cAMP in vivo and in vitro. Therefore, a novel Markov model of PKA-RI[alpha] activation in response to cAMP binding was developed to determine if a more physiological model of PKA activation could explain the observed in vivo PKA activity. We used cAMP-R-subunit binding experiments as well as PKA phosphorylation rates from literature to show that a conformational selection mechanism is the best explanation for PKA activation by cAMP. Furthermore, this model was able to predict cAMP binding to mutant R-subunit with an inhibited A or B binding domain in the absence or presence of C-subunit. Finally, we propose a new model for PKA regulation which incorporates direct binding between PDE and the R-subunit of PKA as a novel pathway to stimulate PDE activity and inhibit PKA activation. We used in vitro kinetic experiments with 3H-tagged cAMP and purified PKA and PDE to quantify this interaction and explore its specificity both for PKA and PDE isoforms. The previous PKA Markov model was adjusted to include these potential states and it was fit to experimental data of PDE activity in the presence of R-subunit. The model was able to fit this data and predict the effect of PDE binding on the PKA activity as well. This work suggests that incorporating this interaction into current whole cell models will significantly improve our understanding of how PKA is activated in viv